The Historical Genesis of Asset Classification: Publix Supermarkets, Inc. v. United States (1992)

The United States Claims Court decided Publix on April 28, 1992. The dispute concerned whether supermarket HVAC systems qualified for investment tax credit treatment or were excluded building structural components. The IRS’s Cost Segregation Audit Techniques Guide identifies the decision as classifying the HVAC system as Section 1250 property. The court was then named the United States Claims Court; its later name, the United States Court of Federal Claims, should not be applied retrospectively without explanation.

The relevant framework was former Section 48 and Treasury Regulation Section 1.48-1. That regulation generally includes central heating and air-conditioning systems among structural components. It also provides an exception for machinery installed solely because temperature or humidity requirements are essential to operating other machinery or processing materials or foodstuffs. Incidental employee comfort does not, by itself, defeat that exception. Accordingly, the governing inquiry cannot accurately be reduced to a rule that any equipment benefiting occupants is automatically disqualified.

Publix’s argument concerned the relationship between supermarket air conditioning and refrigeration. The HVAC systems were held ineligible for the investment credit. The later Ninth Circuit decision in Albertson’s, Inc. v. Commissioner, 38 F.3d 1046 (1994), expressly agreed with Publix’s treatment of HVAC systems. This establishes the investment-credit context; it does not establish a ruling about software debugging, technical experimentation, or Publix’s Section 41 eligibility.

The practical lesson is to identify the specific tax provision and its expense or property definitions. There is no established basis here for asserting that the IRS routinely invokes Publix to reject Section 41 claims or that a building becomes a qualified research expense merely because the taxpayer calls it an experimental business component.

Distinct Frameworks: Investment Credits and Section 41 R&D Credits

The research credit measures eligible research expenditures rather than providing a general credit for purchasing innovative equipment. Its statutory tests and exclusions operate independently of the structural-component rules applied in Publix. A retailer’s investment may improve business operations without satisfying Section 41, while eligible research activities may occur within an ordinary building.

Under Section 41(d), the research must satisfy the applicable research-expenditure requirement, seek technological information useful in developing or improving a business component, and involve a qualifying process of experimentation for a permitted purpose. Following the 2025 amendment, Section 41(d)(1)(A) refers to domestic research or experimental expenditures under Section 174A, generally for taxable years beginning after December 31, 2024. Historical claims must be evaluated under the provisions applicable to their own tax years.

• Research uncertainty: Available information must leave uncertainty about capability, method, or appropriate design in developing or improving the business component.
• Technological basis: The experimental process must fundamentally rely on physical or biological sciences, engineering, or computer science.
• Permitted purpose: The intended improvement must concern function, performance, reliability, or quality, rather than cosmetic, seasonal, style, or taste factors.
• Process of experimentation: The taxpayer must identify uncertainty and alternatives and evaluate those alternatives through an appropriate process, such as modeling, simulation, or systematic trial and error. The regulatory substantially-all requirement generally uses an 80% threshold measured by costs or another consistently applied reasonable basis.

For a hypothetical robotic warehouse project, qualifying employee services, eligible supplies, and eligible contract research may enter the credit computation. Purchasing depreciable robotic equipment does not itself produce a qualified supply expense. Section 41(b)(2)(C) excludes land, improvements to land, and depreciable property from supplies. The absence of general-purpose utility does not override those exclusions. Software development must separately satisfy the research tests and any applicable software rules.

Internal Use Software and the High Threshold of Innovation

Large retailers use software for inventory, logistics, customer services, and administrative functions. Publix’s current corporate facts page states that it employs more than 260,000 people; the supplied study’s 207,000-person figure should not be presented as current. The scale of a company’s operations does not determine whether its software development qualifies for a research credit.

Regulatory Framework for Software Credits

The final internal-use software regulations were issued in 2016. The supplied study does not establish a separate substantive 2021 amendment to that framework. Treasury Regulation Section 1.41-4(c)(6) generally treats software developed primarily for general and administrative functions as internal-use software. These functions include financial management, human resources, and support services.

Software developed to be sold, leased, or licensed to third parties, or to enable qualifying interactions with third parties, is generally outside that definition. Intended use at the beginning of development matters. Mixed-use systems require analysis of their functions and identifiable subsets, and being customer-facing does not automatically establish qualified research.

The dual-function safe harbor permits 25% of otherwise qualified research expenditures for the relevant dual-function software or subset to enter the credit computation when anticipated third-party interaction represents at least 10% of its use and the other conditions are satisfied. It is neither a 25% credit rate nor an automatic exemption for all mixed-use software.

Where the high threshold of innovation applies, the taxpayer must establish all three additional requirements:

• Innovation: Successful development would produce a substantial and economically significant reduction in cost, improvement in speed, or other measurable improvement. Novelty alone is not the test.
• Significant economic risk: Substantial resources are committed, with substantial uncertainty because of technical risk that those resources will be recovered within a reasonable period. Ordinary budget risk or possible litigation does not establish this requirement.
• Commercial availability: The software cannot be purchased, leased, or licensed and used for its intended purpose without modifications that themselves satisfy the innovation and significant-economic-risk requirements. A wholly unique creation is not invariably necessary.

These standards arise from Section 41 and its regulations. They should not be attributed to the Publix HVAC decision.

Judicial Scrutiny: The Phoenix Design Group Ruling (2024)

Phoenix Design Group, Inc. v. Commissioner, T.C. Memo. 2024-113, issued December 23, 2024, directly addressed research credits claimed by an engineering firm. The court found that the three projects tried as a sample did not establish qualified research. Its analysis addressed the research-expenditure uncertainty requirement and the process of experimentation, rather than applying Publix’s asset-classification rules.

The firm’s design process, revisions, professional calculations, and general descriptions of engineering challenges did not establish the necessary qualifying activities for those projects. The decision illustrates the importance of connecting specific technical uncertainties, alternatives, and evaluations to the claimed work. It does not categorically exclude engineering that uses existing scientific principles, building codes, or established tools.

The parties’ stipulation linked accuracy-related penalties to the outcome of the credit dispute. Describing the resulting 20% penalties as a newly announced IRS policy against retrospective studies would overstate the decision. Nor does the case create a universal requirement to use one particular type of contemporaneous record.

Administrative Challenges: Refund-Claim Review and Substantial Variance

Research-credit refund claims must satisfy filing requirements as well as the substantive credit rules. A sufficiently detailed filing does not guarantee allowance, and a substantively supportable credit can still face procedural problems if the refund grounds are inadequately stated.

Research-Credit Refund-Claim Review

The IRS’s published research-credit refund FAQs describe review by campus employees or field examiners, in coordination with subject-matter experts and counsel. They do not substantiate the supplied study’s assertion that an automated “Classifier” rejects weak claims without human review.

Under the IRS’s published guidance, taxpayers must identify the relevant business components and research activities and provide total qualified wage, supply, and contract-research expenses. Effective June 18, 2024, the IRS waived the filing-stage requirement to provide individual researchers’ names and the information each sought to discover, although that information may be requested during examination. The published transition period runs through January 10, 2027, with a 45-day opportunity to perfect deficient claims within that process.

The Substantial Variance Doctrine

Treasury Regulation Section 301.6402-2(b)(1) requires a refund claim to state its grounds and sufficient supporting facts. The substantial variance doctrine generally limits a refund suit to grounds fairly presented in the administrative claim, subject to applicable exceptions and waiver principles. Whether later software or hardware details merely elaborate an existing ground or introduce a new one depends on the actual filing and procedural record.

The supplied study’s Shleifer reference has not been established here as authority for its categorical inventory-software example. A prudent filing identifies the actual components and activities underlying the requested refund rather than relying on a different theory first advanced in court.

Supervisory Approval and Document Integrity

Section 6751(b) generally requires written supervisory approval of the initial determination of covered penalties, subject to statutory exceptions and applicable timing rules. A failure to satisfy an applicable approval requirement may provide a defense to a penalty; it does not establish entitlement to the underlying research credit.

Allegations about summons authorization in a separate individual’s tax proceeding should not be presented as proof of a defect in research-credit penalty approval. The supplied study does not establish that the Jenkins Fancelli matter involved a Publix Section 41 claim or a Section 6751 penalty determination.

Illustrative Case Study: Point of Sale Systems and Technical Innovation

A hypothetical retailer developing a point-of-sale system can illustrate the research-credit analysis. This example does not describe an established Publix research project, credit entitlement, or judicial finding about its checkout systems. Consumer allegations alone would not establish a technical defect, its cause, or qualifying research.

Suppose a retailer investigates unresolved latency or consistency problems in synchronizing prices and transaction data. The relevant question is whether its development activities satisfy Section 41, rather than whether the project is expensive, important, or undertaken in response to a complaint.

• Business component: Identify the actual software or relevant subsystem being developed or improved.
• Experimentation: Identify the technical uncertainty and document evaluation of competing architectures or synchronization methods, including the results and resulting design decisions.
• Software classification: Determine intended functions at the start of development. Checkout software is not automatically internal-use software merely because it handles payments. Customer interaction, administrative functions, and separable subsets require analysis.
• Expense eligibility: Link qualifying services and other eligible expenditures to the research and separate excluded activities and costs.

Routine maintenance, ordinary quality-control testing, and implementation using an already established solution generally do not demonstrate the required experimental process. A new improvement project can nevertheless be evaluated on its own facts. Failure to qualify would not turn software costs into building structural components under the Publix decision.

Financial Context and Tax-Credit Eligibility

A retailer may consider available tax incentives when planning investments. Financial pressure, litigation exposure, or an employee-ownership structure does not demonstrate research-credit eligibility. The supplied study does not substantiate its assertions that unrelated litigation drives Publix’s research-credit strategy or that research-credit savings are directly transferred to its retirement plan.

Publix identifies itself as the largest employee-owned company in the United States. That ownership model supplies business context, but claims about guaranteed retirement benefits from a particular tax credit require separate evidence.

Comparative Jurisprudence: Suder v. Commissioner (2014)

Suder v. Commissioner, T.C. Memo. 2014-201, illustrates the fact-specific evaluation of research activities and supporting evidence. It should not be presented as a blanket lenient exception to Section 41 or as a doctrinal conflict with Phoenix Design Group. Qualification and the amount of eligible expenses remain distinct questions.

The governing regulations do not require an advance in knowledge across the entire industry. A taxpayer may know that an objective is achievable while remaining uncertain about the method or appropriate design. However, an assertion that a better solution might exist is not sufficient without qualifying uncertainty and an experimental process.

Existing manuals, public information, and standard techniques must be assessed in context. Their existence is not an automatic disqualification, but activities limited to duplication, routine testing, or other statutory exclusions remain ineligible. Supported testimony and reasonable estimates may contribute to proof; generalized retrospective assertions do not establish entitlement by themselves.

Implications for R&D Tax Credit Applications

The central task is to substantiate the actual research, eligible expenses, and filing position under the law applicable to the claim year. These requirements arise from Section 41 and related authorities, rather than a new line of R&D precedents descended from Publix.

Applying the Shrinking-Back Rule

Treasury Regulation Section 1.41-4(b)(2) provides for applying the qualification requirements to successively smaller subsets when they are not satisfied for the overall business component, until a qualifying subset or the most basic element is reached. The rule does not permit taxpayers to select costs arbitrarily or disregard other ex